You’re re-stating my (and Krugman’s) point. Insurances have an issue with market failure based on asymmetric information, but, as dsquared points out, that issue only arises as long as the insurances cannot divine a client’s future need for healthcare. But insurances have gotten pretty good at forecasting by now, and to the extent that they can price-discriminate on the expected expenditure stream the health insurance market should approach efficiency for all those who participate in it, at least on that count. (Public health also creates an externality problem, but let’s ignore that for a second.) But such a market will be nowhere near universal, especially because the fee structure doesn’t come close to meeting our universal affordability goal. Krugman makes exactly that point but Tabarrok in typical econbot mode cannot distinguish between this kind of failure and the textbook efficiency-driven market failure, hence his ongoing argument that oh but health insurance could be efficient if’n for all that regulation. Except it would be like health insurance for pets: a luxury good for those who can afford it, with the alternative of euthanasia for those who can’t.

But those goals aren’t ones that a market can sensibly be expected to meet. A market only serves those who come to market with something to trade, and it only serves them to the extent of the value they bring to it. If you want some members of society to be able to receive health care whose market value exceeds their economic productivity, you want something that no market can possibly provide because providing that thing is not the sort of thing that markets do. The market solution to a poor person with cancer is euthanasia – the cost of keeping them alive exceeds the value of doing so.

You can reduce the need for outright redistribution to some extent by describing health care as a matter of insurance and paying today for a right to treatment tomorrow, hiding the division between people who need expensive treatment and people who don’t behind a veil of ignorance. (They all pay, and then find out who needs more treatment and who needs less.) But insurance companies have proven rather good at tearing apart this veil and separating the different groups from each other to the extent that many members of the high-risk group can’t afford to pay for even the actuarial prediction of how much health care they will need. (They’ve also gotten rather good at selling the expectation of coverage without, technically, in the fine print, selling the actual coverage. This is obviously profitable, but I don’t think anyone even on the right would describe it as socially desirable.)

Ultimately, either you believe that all members of society have a moral right to draw on the public purse (i.e. draw on all other members of society, willing and otherwise, via the tax mechanism) for some level of legitimate health care expenses, or you don’t. In the former case, markets are obviously going to be inadequate to distribute the massive financial burden of very bad health events over broader shoulders than the patient’s (since that goal cannot be accomplished by an exchange of equal values), and in the latter, markets can do just fine at providing individuals with however much health care they can individually afford, so if you think that’s all they should get, then there is no problem.

Tabarrok, oddly for an otherwise intelligent person who is presumably aware that Krugman is a liberal, doesn’t perceive the essentially normative disagreement underlying the different points of view he and Krugman bring to the problem. Whether markets are good enough at delivering health care depends greatly on how you measure good, and how good is good enough.

Re #34, the asymmetry exists if insurers price-discriminate by demographic factors. If you belong to a demographic that, according to the insurance’s reckoning, will require high levels of healthcare due to poor habits, it doesn’t matter that you yourself follow a healthy lifestyle. You will still be charged at the poor-habits price. That’s a simple lemons market.

Of course Tabarrok also misinterprets Krugman’s claim that “health care can’t be marketed like bread or TVs” as a story of market failure in health insurance. Healthcare is not health insurance, and the inability to reach a social goal with purely private ordering is quite distinct from market failure. We don’t consider the legal system a market failure just because we decided it is better run as a public institution, but we would consider an efficiently run health insurance market a social failure if it doesn’t approach the social goals of universal coverage and affordability.

Finally, the term “adverse selection” refers to a problem of information asymmetry. When talking about an insurer’s willingness to provide insurance to the already sick, no such asymmetry exists.

And that’s the intelligent part of his post. Before he starts ignoring the little fact that ceteris paribus those who suffer from poor health will demand more healthcare than those in excellent health. (That’s what the word “need” in “likely to need care” means. Need, when endowed with purchasing power, creates demand. Econ 1.) So not only doesn’t he understand conditional probabilities, can’t do the numbers (21% of those in poor health uninsured vs. 15% in the general population makes for a 40% difference), he also can’t apply simple supply and demand to realize that absent major differences in purchasing power by health class (refuted in Table 15) the only explanation why we see significant less insurance among the poor/fair category than in the general population is that insurers price them out of the market. Iow, they “try to avoid” them.

Embarrassing.

Industrial-strength recision is not a thing forced on insurers by the state. Nothing in the structure of insurance regulations compels firms to pursue every last opportunity to deny every scrap of coverage or simply negate as many policies as possible. (Someone’s going to bring up the whole “fiduciary responsibility” thing. Just for starters, a business that really felt it was being constrained unwisely by that could challenge it in court or seek legislation or just make a PR push, even while abiding by it in the short term. But businesses use it as cover, not as something they actually have any problem with.) And it’s in the area of greatest discretion that we find insurers trying to make most money by providing least coverage.

So the advocates of freeing insurers have, at a minimum, an obligation to demonstrate why these firms, that ones that do exist and would be selling policies in a freer market, would suddenly retract the views they’ve been expressing in testimony to Congress. They say they want to keep screwing us. I believe them. People who don’t need to explain what they know that the insurers themselves either don’t know or are hiding.

Instead, we’re going to get a really poorly-designed public health care system, which will be substantially worse than what we’ve got. We’re going to keep the system that we have, subsidized a little more around the edges, with heavily-regulated-and-mandated employer-connected insurance as the continued norm. Efficiency and monetary savings are extremely unlikely, as every CBO and OMB study has hammered away at.

If I thought a good public system was in the offering, I think it might be worth supporting it, in order to stave off the bad. But if the choice is between the current system and a bad public one, well, that choice is ridiculously easy.

______________
First, this is junky data, although you’d have to be pretty deep into health care wonkery to know that. The CPS is not a health care survey, even though it’s frequently used because it’s easy access and big enough to give state results. The question asking for self-reported “health status—excellent/good-fair/poor” is not a precise indicator of health status. (The insurance questions on the CPS are also way too crude and give answers that are significantly off from almost every specialized health care survey).

But anyway, taking the data at face value, just look at the actual numbers in the chart. Almost 70 percent of the total population says they are in excellent/very good health. So taking yourself out of the top “excellent/very good” category is an indicator that you are in poor health—in the bottom 30 percent of the non-elderly population in terms of health status. These are the people who cost money (the healthiest 50-60 percent of the nonelderly population has basically no costs).

This bottom 30 percent are a lot, not a little, more likely to be uninsured—specifically, the uninsurance rate among the bottom 30 percent in self-identified health status is 18.3 percent, while it’s 13 percent among the top 70 percent. That means that people in the bottom two categories for health status are a whopping 40 percent more likely to be uninsured than people in the most healthy 70 percent! Not a small difference.

Megan obscures this by just glancing at the fair/poor category, where the populations are so small that the absolute percentage point differences don’t look so great. The true percentage differences are a little smaller when you look at only this category, but not that much smaller—the “fair/poor” people have an uninsurance rate of 17.4 percent, which is still fully one-third higher than the healthy folks. (The slight difference may be because of William’s point if you’re actively sick, which this category probably is, you get signed up for Medicaid at the hospital).

Then add on to that that these differences have to be adjusted for the large chunk of people who work at big/prosperous or public sector firms and therefore can get covered by good insurance even when unhealthy. My guess is that the gap would be much greater among those working for smaller firms or the self-employed. You also have to adjust for the flaws in using a 1/0 indicator of insurance, which is all that the CPS gives you. What is needed is a measure of real coverage, not insurance status. The individual insurance market is notorious for selling policies that count as “insurance” but don’t really cover you when you get sick, because of pre-existing condition clauses, annual or treatment limits, or a refusal to renew when you get sick. This is well known in health policy circles.

Anyway, all of this is conflated about nine different ways because income and health status are so deeply connected, and health status also gives you a greater incentive to sign up for insurance. So you have biases going both ways. But there is basically no evidence in this simple tabulation that Krugman is wrong.

First, his gross logical error of reasoning with a converse: If insurance companies do avoid covering people who are “likely to need care,” this suggests that the uninsured are unhealthy. Numerous people have shown by real world example why this is unsound.

But the second, and far worse problem is the libertarian ideological viewpoint that won’t let him understand Krugman’s liberal statement.

As a liberal, Krugman looks at the desired results: high quality health care for the nation as a whole. The kind of statistics that would let us compete with the other industrialized nations in terms of costs and outcomes. As a liberal, Krugman is willing to employ diverse methods to achieve this goal: whatever tools will work best singly or in combination. And if a solution isn’t working well, he’s willing to try another. This is the classical liberal pragmatism that has been employed in the US since even before its founding.

As a libertarian ideologue, Tabarrok cares less about what the result is, and more about how the result is achieved. Hence his incessant focus on markets and rejection of government. A perfectly efficient market result could have radically different health results for rich and poor, and libertarians would find little reason to criticize it. Markets enfranchise one dollar as one vote, which grossly disenfranchises the poor compared to the liberal one person one vote system.

Libertarians confuse the means (markets) with the desired ends (in this case universal health of the populace—not any particular health care program.) If we let their hired propagandists keep misdirecting us away from what we want (good health care for all) with incessant claims that markets will give us everything we want AND A PONY, then we fall prey to the great conservative goal: preserving the rights of people who own property at the cost of decent lives for everyone else in the world.

More importantly, the use of stats is pretty suspect.

McArdle says the percentage of the uninsured* who are in “fair” or “poor” health really isn’t much larger than the percentage of the insured in those categories: 10.3% of the uninsured, versus 8.4% of the insured. [the asterisk appears to lead nowhere].
That’s just wrong (or rather not known to be right). The 8.4 figure refers to the total nonelderly adult population, including both insured and uninsured – and in particular those on public insurance who are by far the least healthy according to these data. (A bit of simple arithmetic – using the rounded data – gives a figure of 7.99% for ‘all insured’ – but McArdle clearly didn’t know that).

And Tabarrok gives us – 60% of the uninsured are in excellent health as though that alone tells us anything useful. Not much comfort for the other 20m people, is it. And it’s not ‘excellent’, but ‘excellent or very good’ – the latter presumably being significantly unhealthier.

I’m not sure that table 10, even with the arithmetic which giving Tabarrok the benefit of the doubt we assume was applied – really shows that overall the uninsured are only slightly less healthy than the insured, either. It certainly can do so only given some clear idea of what counts as ‘slightly’. And even then, why are the publicly insured being counted with the private?

(As an aside, a survey asking for subjective assessment of one’s own health (if that’s what was done) has some obvious problems relating to expectations and adaptation levels, which may or may not have been corrected for somehow, but I suppose one takes the rough with the smooth in preference to eternal stat wars?)

And if any conclusions are supposed to be derived from such data, then among other factors like traditional adverse selection, the effects on health of having had insurance in the past would have to be factored in too.

Anyway, a quick bit of sheetspreading (and a slightly-slower-than-I-had-bargained-for bit of formatting) produced this derived data, probably more useful for these purposes despite some rounding issues:


. . . . . . .employer . indiv . . .m/aid. . . other. . . none
. . . . . . . . . . . . . . . . . . . . . . . public
.

% dist by coverage type
. . . . . . . . . . . .
Exc/VG . . . 69.80% . . 6.80% . . . 4.50% . . 1.50% . . 17.40%
Good . . . . 58.20% . . 4.80% . . . 9.10% . . 3.20% . . 24.80%
Fair/Poor. . 36.30% . . 3.70% . . .25.80% . . 11.90% . . 22.20%
.

.
millions
. . . . . . . . . . . .
Exc/VG . . 126.8266 . .12.3556 . . 8.1765 . . 2.7255 . . 31.6158
Good . . . .33.6396 . . 2.7744 . . 5.2598 . . 1.8496 . . 14.3344
Fair/Poor. . 7.9497 . . 0.8103 . . 5.6502 . . 2.6061 . . .4.8618
All . . . .168.4159 . .15.9403 . .19.0865 . . 7.1812 . . 50.8120
.

.
% dist by health status
. . . . . . . . . . . .
Exc/VG . . . 75.31% . . .77.51% .. 42.84% . . 37.95% . . 62.22%
Good . . . . 19.97% . . .17.40% .. 27.56% . . 25.76% . . 28.21%
Fair/Poor . . 4.72% . . . 5.08% .. 29.60% . . 36.29% . . .9.57%
All . . . . 100.00% . . 100.00% . 100.00% . .100.00% . .100.00%

Perhaps more useful groupings below. I suggest looking at the third block and comparing ‘private’ to ‘none’ (or even to ‘public + none’), bearing in mind that ‘Good’ presumably actually stands for ‘moderately unhealthy’, as well as potential self-assessment issues mentioned above:


. . . . . . .all. . . . all . . . private. . public. . . none
. . . . . . . . . . . . insured . . . . . . + none
.

% dist by coverage type
. . . . . . . . . . . .
Exc/VG . . . 100.00%. .82.60%. . 78.10%. . . 23.40%. . .17.40%
Good . . . . 100.10%. .75.30%. . 66.20%. . . 37.10%. . .24.80%
Fair/Poor. . .99.90%. . 77.70%. . 51.90%. . . 59.90%. . .22.20%
.

.
millions
. . . . . . . . . . . .
Exc/VG . . . 181.7. . .150.0842. .141.9077. . 42.5178. . 31.6158
Good . . . . .57.8. . . 43.5234. . 38.2636. . 21.4438. . 14.3344
Fair/Poor. . .21.9. . . 17.0163. . 11.3661. . 13.1181. . .4.8618
All . . . . .261.4. . .210.6239. .191.5374. . 77.0797. . 50.8120
.

.
% dist by health status
. . . . . . . . . . . .
Exc/VG . . . .69.51%. . 71.26%. . . 74.09%. . .55.16%. . 62.22%
Good . . . . .22.11%. . 20.66%. . . 19.98%. . .27.82%. . 28.21%
Fair/Poor . . .8.38%. . .8.08%. . . .5.93%. . .17.02%. . .9.57%
All . . . . .100.00%. .100.00%. . .100.00%. . 100.00%. .100.00%

(BTW the word ‘spreadsheeting’ does not, contrary to initial perception, occur above.)

I.e., not that the uninsured are disproportionately unhealthy, but that the unhealthy are disproportionately uninsured.

Moreover, the large number of healthy, uninsured-by-choice isn’t a refutation of Krugman but the flip-side of his argument.

The latter does not imply that “more markets” will solve the problem. If the market failure is due to the high cost of market-provided insurance, as it seems to be, then “more markets” is not likely to solve the problem alone. By the same token, any government-based solution must confront the same problem: “mandating” coverage will not help the uninsured purchase coverage. Generous subsidies would. But what is driving much of the debate is the cost of medical insurance, not its extent. And that’s the underlying problem: we’re having a healthcare debate in the US because its expensive, not because care is inadequate.